Sora is a generative AI platform that creates great-looking videos from just text

Competition in the generative AI space has been heating up. And after launching the likes of ChatGPT and Dall-E 3, OpenAI has just announced what could be the next big thing in this space. Enter Sora, a new generative AI platform that can create videos from text prompts. But what’s so good about it you ask? It’s the sheer quality of videos that are being created from scratch.

Sora: How the platform works and what it can do

Sora allows users to create life-like minute-long videos based on textual prompts. As per OpenAI, it is capable of creating extremely detailed videos with several shots – featuring multiple characters, detailed background and foreground elements, motion patterns, and more.

The model is currently available for testing only to a handful of filmmakers, designers, and content creators, for feedback on the model can be improved upon, before its rollout to the general public.

ALSO READ: ChatGPT is getting human-like memory; will soon remember you and things you like

This is because the existing model, while fairly capable, does have a few shortcomings. OpenAI has accepted that the model currently struggles to simulate physics and doesn’t entirely understand cause and effect.

It gives the example of a person taking a bite out of a coolie in one such video which was generated using the platform, wherein the cookie appeared to be untouched after the person had put it back down.

Sora: Safety mechanisms to prevent potential misuse

OpenAI says the platform has been developed within the C2PA guidelines. C2PA, for the unaware, stands for Coalition for Content Provenance and Authenticity. The organisation drafts and enforces technical standards for unique content, to ensure its authenticity and protect It from potential misuse.

ALSO READ: Here’s an AI tool that will protect your online images from manipulation

As such then, all videos generated on Sora will have a watermark similar to DALL-E, to prevent misuse, and neither will the platform allow users to create videos based on text prompts that could potentially be harmful or offensive to anyone’s social or religious beliefs.

When will Sora be made available for use to the public?

As mentioned above, Sora has currently only been available for testing to a handful of filmmakers and content creators. When the model will be rolled out for users globally, and whether it will be available only on laptops and desktops, or smartphones as well, remains unknown as of now.